A fault prediction system and method based on a Bayesian belief network

Abstract

The invention provides a fault prediction system and method based on a Bayesian belief network, and the method comprises the steps of building a system mathematical model based on the Bayesian beliefnetwork according to a system structure and internal connection, wherein the model comprises the parameters consisting of a system hierarchical interconnection structure and a condition probability; calculating system failure rate analysis weak links through a Bayesian belief network reasoning engine according to the failure rate of root node components; and according to a reliability correlationformula, predicting the average time before failure of the system. According to the fault prediction system and method based on the Bayesian belief network provided by the invention, for the system with high uncertainty and complex internal interconnection relationship, reliability analysis can be more accurately and quickly carried out on the system, weak links of the system can be found, the fault rate of the system can be calculated, and the service life of the system can be predicted, so that the engineering application value is better.

Description

technical field [0001] The invention relates to a fault prediction system and method based on a Bayesian belief network, belonging to the technical field of fault diagnosis and prediction. Background technique [0002] There are many methods of fault diagnosis and prediction, but the determination and selection of fault diagnosis and prediction methods involve many factors, such as the type of failure, the form of failure mode, the criterion of failure, the method of processing data, and the uncertainty of variables. Considerations and so on, which bring huge problems to the practical application of fault diagnosis and prediction. [0003] Bayesian belief network has many advantages as a modeling analysis method based on statistics. It uses a mathematical model to quantitatively describe the causal relationship and degree of correlation between variables in the form of probability. Moreover, the reasoning of Bayesian belief network is bidirectional, both bottom-up and top-...

AI Technical Summary

Problems solved by technology

[0002] There are many methods of fault diagnosis and prediction, but the determination and selection of fault diagnosis and prediction methods involve many factors, such as the type of failure, the form of failure mode, the criterion of failure, the method of processing data, and the uncertainty of variables. considerations, etc., which pose a huge problem for the practical application of fault diagnosis and prediction

Images